CN105962930A - Learning state sectional-type recording method and device, and learning state sectional-type displaying device - Google Patents
Learning state sectional-type recording method and device, and learning state sectional-type displaying device Download PDFInfo
- Publication number
- CN105962930A CN105962930A CN201610119931.2A CN201610119931A CN105962930A CN 105962930 A CN105962930 A CN 105962930A CN 201610119931 A CN201610119931 A CN 201610119931A CN 105962930 A CN105962930 A CN 105962930A
- Authority
- CN
- China
- Prior art keywords
- learning state
- section
- data
- value
- state data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/369—Electroencephalography [EEG]
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
- A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
- A61B5/1103—Detecting eye twinkling
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61F—FILTERS IMPLANTABLE INTO BLOOD VESSELS; PROSTHESES; DEVICES PROVIDING PATENCY TO, OR PREVENTING COLLAPSING OF, TUBULAR STRUCTURES OF THE BODY, e.g. STENTS; ORTHOPAEDIC, NURSING OR CONTRACEPTIVE DEVICES; FOMENTATION; TREATMENT OR PROTECTION OF EYES OR EARS; BANDAGES, DRESSINGS OR ABSORBENT PADS; FIRST-AID KITS
- A61F5/00—Orthopaedic methods or devices for non-surgical treatment of bones or joints; Nursing devices; Anti-rape devices
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Veterinary Medicine (AREA)
- Animal Behavior & Ethology (AREA)
- Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Public Health (AREA)
- General Health & Medical Sciences (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Physics & Mathematics (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Surgery (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
- Dentistry (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Ophthalmology & Optometry (AREA)
- Nursing (AREA)
- Orthopedic Medicine & Surgery (AREA)
- Vascular Medicine (AREA)
- Psychiatry (AREA)
- Psychology (AREA)
- Physiology (AREA)
Abstract
The invention discloses a learning state sectional-type recording method and device, and a learning state sectional-type displaying device. The learning state sectional-type recording method comprises the following main steps: carrying out segmentation on learning states according to set conditions by adopting a processor; extracting a continuous value reflecting the length or time frame of the segment of state data, or extracting a characteristic value reflecting the size or state of the segment of state data; and recording the characteristic value and the continuous value of the segment of learning state data. With the adoption of the method and the devices, the learning state of students can be monitored, the state data in the learning process is recorded, by analyzing the recorded learning state data, the bad learning habits or problems of the students can be found, and the students grow healthily; through the learning state data recorded in a sectional-type manner, the format is simple, the data size is small, the requirements on hardware are lowered, the follow-up processing and analysis are easy, and a statistical chart drawn according to the characteristic value and the continuous value can directly and clearly reflect the distribution characteristics of the learning state data.
Description
Technical field
The present invention relates to electronic technology, particularly a kind of learning state fragmented recording strategy and record and display device.
Background technology
Children's control is relatively poor, in learning process, some bad habits often occurs, such as gazes around, absent-minded, abnormal sitting posture, during reading writing with eye distance close to too etc..If Students ' Learning situation to be carried out record, the problem that reflection exists, correct in time, can effectively help the growth of children.Wearable device is limited by volume weight, and storage capacity is limited, simplifies the record of data, beneficially wearable device experience, and follow-up Data Management Analysis.
Summary of the invention
First purpose of the present invention is to provide a kind of learning state fragmented recording strategy, learning state data are divided into some the sections comprising a learning state data feature values and a data length persistent value, the data making record are easily understood, it is simple to follow-up Treatment Analysis.
Second object of the present invention is according to above-mentioned fragmented recording strategy, it is provided that a kind of learning state segmentation recording equipment, reduces learning state data recorded amounts, saves memory space.
Third object of the present invention is according to above-mentioned fragmented recording strategy, it is provided that a kind of learning state segment display device, succinct display learning state distribution characteristics directly perceived.
First technical scheme of the present invention, a kind of learning state fragmented recording strategy, key step includes that processor carries out segmentation according to imposing a condition to learning state；Extract this section of status data length of reflection or the persistent value of period, or extract this section of status data size of reflection or the eigenvalue of state；The eigenvalue of minute book section learning state data and persistent value.Described learning state includes head pose, by least one in, brain states of eye distance.The present invention is the status data in a complicated learning process, it is divided into some periods or section, the eigenvalue further extracting described period or sector data carries out record with persistent value, on the one hand by eigenvalue reflection learning state size of data or state, on the other hand by persistent value reflection data length or period, record data are simple, directly perceived, understandable, and data volume is little.
Described segmentation method can according to single learning state data with impose a condition relatively carry out segmentation, or according to multiple learning state data with impose a condition relatively carry out segmentation；Single status data are divided by real time, it is considered to the individual character of each data, and the latter divides according to batch of data characteristic, the main general character reflecting this batch data.Can be continuous print between each segment data, it is also possible to be discontinuous, between section and section continuously, make segmentation limit cover total data；Being interrupted if existed between section and section, record main period of time or sector data, being interrupted and omitting of short duration abnormal data, with strong points.
Second technical scheme of the present invention, a kind of learning state data sectional recording equipment, including learning state sensor, processor and memorizer, wherein the transport processor calculating of learning state sensor acquisition learning state signal processes, generate learning state data, and according to above-mentioned fragmented recording strategy, learning state data are carried out segmentation, extract the persistent value of this section of status data length of reflection, or extract the eigenvalue of this section of status data size of reflection；The eigenvalue of memorizer minute book section learning state data and persistent value.Reflect each section of learning state with eigenvalue and persistent value, save memory space, and be easy to subsequent treatment.
Described learning state sensor includes head pose sensor, uses at least one in eye range sensor, brain electric transducer, and wherein said head pose sensor includes the one in Gravity accelerometer, gyroscope, magnetometer, eye tracker；Described eye range sensor includes the one in infrared range-measurement system, ultrasonic range finder, laser range finder；Described brain electric transducer includes single channel brain electric transducer or multichannel brain electric transducer.
Further improving, recording equipment can arrange data transmission interface, by coffret, record data is sent to external equipment, and described data transmission interface is wired or wireless mode；Wired mode includes RS232 or USB interface, and wireless mode includes any one in bluetooth and radio frequency and zigbee and wifi technology.
Further improving, recording equipment can arrange prompting module, reminds section when described learning state data fall into setting, and processor triggers prompting module and reminds, and described prompting module includes any one in sound, light, micro-electrical stimulation, vibrations, bone conduction.
3rd technical scheme of the present invention, a kind of learning state data sectional display device, for the learning state data of record are carried out segment processing, show over the display, including the communicator receiving learning state data, data are carried out processor and the display of display statistical result of segmentation statistics, it is characterized in that processor carries out segmentation according to above-mentioned segmentation method to learning state data, the persistent value of statistics reflection this segment data length, draw the eigenvalue of each section of learning state data and the statistical graph of persistent value, transmission display shows, the distribution characteristics of learning state data is intuitively reflected by statistical graph.
Students ' Learning state is monitored by the present invention, the status data during recording learning, by being analyzed the learning state data of record, it appeared that the bad study habit of student or problem, helps it to grow up healthy and sound.By the learning state data of segmentation record, form is simple, and data volume is little, reduces the requirement to hardware, and is prone to subsequent treatment analysis, and the statistical graph drawn according to eigenvalue and persistent value can reflect learning state data distribution characteristics simple and clearly.
Accompanying drawing explanation
Fig. 1 is the configuration block diagram of segmentation recording equipment embodiment.
Fig. 2 is head pose sensor circuit catenation principle schematic diagram.
Fig. 3 is from sensor circuit catenation principle schematic diagram with eye distance.
Fig. 4 is brain electric transducer circuit catenation principle schematic diagram.
Fig. 5 is segmentation record basic flow sheet.
Fig. 6 is the configuration block diagram of display device embodiment.
Fig. 7 is a mental status statistic histogram example.
Detailed description of the invention
In the recording equipment shown in Fig. 1 configures block diagram, including learning state sensor, processor and memorizer, wherein learning state sensor acquisition learning state signal transport processor, generate learning state data, and learning state data are carried out segment processing, transmit memorizer record and preserve.Learning state sensor includes head pose sensor, uses at least one in eye range sensor, brain electric transducer.Below in conjunction with in particular state sensor circuit connection diagram, it is shown that communication connection, do not show decoupling and all connections, processing controller ADuC7024 has memory element, being programmed by UART, SW1 is on and off switch, SW2 and SW3 is to reset and lower load switch respectively.
Head pose includes head inclination angle, face orientation, eye gaze direction etc., attitude transducer includes Gravity accelerometer, gyroscope, magnetometer, eye tracker, the circuit theory schematic diagram of acceleration transducer is used shown in Fig. 2, processing controller ADuC7024 and numeral Gravity accelerometer ADXL345 is communicated by 4 wire types SPI, and pin 8,9 is to interrupt control.Acceleration transducer ADXL345 is placed in head, it is used for monitoring head inclination angle, when head run-off the straight, gravity changes in the weight component output signal that X, Y, Z tri-is axial, the size of output is the most relevant with the angle of vertical direction with 3, and the least with the angle of vertical direction, its output is the biggest, otherwise, export the least.Processing controller ADuC7024 reads the weight component size of tri-axial outputs of acceleration transducer ADXL345, calculates 3 axial and vertical direction angles, thus calculates head obliquity information.
At Fig. 3 eye distance in the circuit theory schematic diagram of record, using infrared distance sensor GP2D12, its investigative range 10 ~ 80cm, the analog voltage of corresponding output 2.55 ~ 0.42V, range finding is inversely proportional to voltage.The output end vo ut of the analog input end ADC4 and infrared distance sensor GP2D12 of processing controller ADuC7024 is connected, and reads the analogue signal of infrared distance sensor GP2D12 output.The analog voltage of infrared distance sensor GP2D12 input, treated controller ADuC7024 internal conversion becomes digital voltage.
Brain electric transducer includes single channel or multichannel, uses one pole or bipolar protocol to gather EEG signals；Described single channel is the EEG signals only monitoring one region such as forehead of head, is single channel as god reads science and technology ThingkGear AM chip；Described multichannel is the EEG signals in multiple regions such as monitoring head such as forehead, the crown, rear pillow, if Dezhou ADS1299 chip is 8 passages.The faint instability of EEG signals, is disturbed by strong background noise again, and brain electric transducer is through the noise reduction process strengthened EEG signals and to background noise, and each wave band EEG signals of output can be used directly to do further applied analysis.Shown in Fig. 4 is single channel EEG brain wave acquisition sensor ThingkGear AM family chip circuit catenation principle schematic diagram, processor ADuC7024 port P1.5 is connected with brain electric transducer ThinkGear AM input RXD, for brain electric transducer ThinkGear AM carries out the operations such as initialization.Tri-metal electrodes of A, B, the C contacted with human brain respectively with the acquisition electrode EEG of brain electric transducer, compare electrode REF and earth terminal GND and be connected, 512 EEG signals data points of brain electric transducer collection per second, extract the EEG signals of eight wave bands (Delta, Theta, LowAlpha, HighAlpha, LowBeta, HighBeta, LowGamma, MiddleGamma), with three eSense parameters: focus, allowance and nictation are detected, and are exported by port TXD.The brain states that the reflection of different-waveband EEG signals is different, such as Delta(0.5Hz ~ 3Hz) ripple reflection be sleep state, also known as " sound sleep ripple ", Theta(3Hz ~ 7Hz) ripple reflection be sleepy state, also known as " shallow sleep ripple ", Alpha(7Hz ~ 13Hz) ripple reflection be light state, also known as " loosening ripple ", Beta(13Hz ~ 30Hz) ripple reflection be consciousness active state, also known as " excitation wave "
Gamma(30 Hz ~ 50 Hz) ripple reflection be tense situation, also known as " pressure " ripple.By neural algorithm based on experimental data base, the power spectrum data using above-mentioned wave band is analyzed calculating, the different mental status can be reflected, as fallen asleep, sleepy, tired, loosen, meditate, be absorbed in, think deeply, anxiety, pressure, excitement, anxiety, like, glad, dejected etc..The index of reflection mental state level can be calculated according to frequency spectrum data, the most above-mentioned focus and allowance are through EEG signals and calculate the mental status index generated, be scope 0 100 numerical value, absorbed angle value is the biggest, illustrate that focus is the highest, otherwise focus is the lowest；Loosen angle value the biggest, illustrate more to loosen, otherwise, illustrate that allowance is the lowest；It addition, mental status criterion can also be set up according to frequency spectrum data, thus mark off the different mental status.Electroencephalogramsignal signal analysis method has: 1. time-domain analysis, the geometric properties of Main Analysis EEG waveform, such as amplitude, average, variance, skewed degree, kurtosis etc.；2. frequency-domain analysis, is analyzed mainly by power spectrum, such as power spectral analysis, coherence analysis etc.；3. time frequency analysis, combines time and frequency and processes, such as the matched jamming analysis etc. to sleep spindle.
In the fragmented recording strategy basic step flow chart shown in Fig. 5, first fragmentation value is set, by the comparison with fragmentation value, it is judged that the section that learning state data fall into, specifically comprises the following steps that
<step 1>start carries out system initialization, enters next step；
<step 2>arranges fragmentation value according to segmentation method；
<step 3>gathers learning state signal, and transport processor generates learning state data, enters next step；
<step 4>judges current affiliated section, enters next step；
Whether<step 5>changes, if it is, show to proceed to new segmentation belonging to judging currently, enter next step, if it does not, extract this section of status data length of reflection or the persistent value of period, or extract this section of status data size of reflection or the eigenvalue of state, proceed to step 3；
The eigenvalue of<step 6>minute book section learning state data and persistent value, proceed to step 3.
Above-mentioned segmentation method includes learning state data are carried out section partition, the learning state data of same section is divided into one section, or learning state data are carried out Time segments division, one continuous time section data be divided into one section.Described learning state data include head pose, with eye distance in, brain states at least one, wherein brain states includes each wave band EEG signals or the mental status calculated according to EEG signals, and described EEG signals includes frequency, power, amplitude, power spectrum, power spectrum.Described persistent value includes that the cumulative frequencies of described section of learning state data or cumulative frequency or persistent period length or persistent period length account for any one in the period of the ratio of total time length or this period of time started and end time or described period；Described eigenvalue includes any one in described section of learning state data sectional value or described section of learning state data statistics value or described section of mental status classification.It is illustrated below in conjunction with specific embodiment.
1, section partition
Section partition is carried out according to learning state data, being the numerical values recited according to learning state data, carry out segmentation according to the fragmentation value set, fragmentation value immobilizes as eigenvalue, even the statistical value of this segment data is as eigenvalue, the change of statistical value is also in the range of fragmentation value, and therefore, every section of learning state size of data or state according to section partition are the clearest and the most definite, have only to extract the persistent value of every segment data, for brain states, can be associated with the mental status.Section partition includes arranging fragmentation value according to learning state data, and the learning state data of same section are divided into one section；Or classify according to the mental status in learning state data, the learning state data under the same class mental status are divided into one section.
(1) fragmentation value is set according to learning state data, fragmentation value is set including any one numerical value in the size according to learning state data or waveform parameter, the learning state data of same section are divided into one section.
It is packet that size according to learning state data arranges a preference of fragmentation value, sets fragmentation value according to the magnitude range of learning state data, the data falling into same piecewise interval are classified as one group, packet the group such as can be away from or different away from packet.
First illustrating brain states data, as a example by the more focus in the brain mental status, focus is the numerical range of 0 ~ 100, can equidistantly be divided into five groups, group away from for 20, the most corresponding sleepy, loosen, tranquil, be absorbed in, excited five kinds of mental status.With [0,20) as a example by section, (" [" expression is equal to, ") " statement), when focus numerical value falls into this interval less than 20, show that current spirit is sleepy, add up persistent value, once focus numerical value is more than or equal to 20, showing to fall into next group, this segment data terminates, and starts to add up the persistent value of next group.Persistent value includes that cumulative frequencies or persistent period are long, if acquisition interval is constant, the product of cumulative frequencies and acquisition interval is exactly the duration of this segment data.If acquisition interval is change, the time separation that acquisition time is two segment records with latest data, or the latter deducts an acquisition time and is spaced as time separation, record time last time separation is as this period of time started, and this separation is as the end time；Or the difference of two time points of record, as the duration of this segment data.Eigenvalue can be any one in the packet higher limit or lower limit or midvalue of class set, or calculate the statistical value of this section of learning state data, statistical value includes any one in meansigma methods, standard deviation, median, mode, or using above-mentioned five kinds of mental status as eigenvalue, the described mental status is replaced the most respectively by numeral 1,2,3,4,5, or by the corresponding described state of kunjuan, fangsong, pingjing, zhuangzhu, xingfen, persistent value record under the corresponding mental status.
For another example as a example by the wave power modal data of Beta in brain states, its magnitude range is 0 ~ 20(× 105μν2), 5,10,15 3 fragmentation values are set, sleep, four kinds of mental status sleepy, clear-headed, excited can be divided, using the Beta wave power modal data statistical value under this state as eigenvalue, or using section value as eigenvalue, adding up the persistent value under every kind of mental status, record is under this section of eigenvalue；Or using the mental status as eigenvalue, replace corresponding state by numeral 1,2,3,4 the most respectively, or with ruishui, kunjuan, qingxing, xingfen correspondence corresponding state, persistent value record under the described mental status.
By same method, head inclination angle being carried out segmentation, with the orthogonal plane in two, left and right before and after first head inclination angle being divided into, rectifying angle when looking squarely is 0 degree, bows as just, swing back as negative in anterior-posterior plane, and Main change scope is-500~500Between；With "Left"-deviationist for just in left and right plane, Right deviation is negative, and Main change scope is-300~300Between, respectively with group away from 50Equidistantly it is grouped, adds up the persistent value under often organizing, carry out record using class boundary value or midvalue of class as eigenvalue.
To with eye distance from for, voltage according to GP2D12 and distance curve of output, between 10-50cm, corresponding voltage data are 2.55 ~ 0.61v, if every 5cm is one group, it is divided into 8 groups i.e. Ai ＜ Ni≤Bi(i=1, 2, 3 ... 8): 2.55v ＜ N1≤1.79v, 1.79v ＜ N2≤1.40v, 1.40v ＜ N3≤1.18v, 1.18v ＜ N4≤0.99v, 0.99v ＜ N5≤0.88v, 0.88v ＜ N6≤0.78v, 0.78v ＜ N7≤0.69v, 0.69v ＜ N8≤0.61v, add up the persistent value under often organizing, record is carried out as eigenvalue using class boundary value or midvalue of class.
The form of grouped record has statistics form and flowing water form two kinds, and wherein statistics form includes that following two form, a kind of form are:
Grouping feature values 1, this section of persistent value, this section of persistent value, this section of persistent value ...
Grouping feature values 2, this section of persistent value, this section of persistent value, this section of persistent value ...
……
Grouping feature values 5, this section of persistent value, this section of persistent value, this section of persistent value ...
Grouping feature values is the fragmentation value or section intermediate value set, and immobilizes, and only records the persistent value of every segment data change, and record format is succinct, is easy to again follow-up data and processes.If recording the time started of every segment data further, in conjunction with persistent value, it may be determined that the period of every segment data, and then whole learning process can be reproduced.
Another kind of form is:
Grouping feature values 1, this group adds up persistent value
Grouping feature values 2, this group adds up persistent value
……
Grouping feature values 5, this group adds up persistent value
The persistent value relation of two kinds of record formats is: this group of grouping feature values 1 adds up persistent value=∑ (this section of persistent value of grouping feature values 1).Former form is in units of section, and record is the persistence length of every section, and the latter is in units of group, and record is the accumulative total persistence length value of this group, and this value is constantly refreshed by statistics in recording process.Persistent value includes cumulative frequencies or persistent period length or this period of time started and end time, in latter record, it is also possible to the cumulative frequency of minute book group or persistent period length account for the ratio of total time length as persistent value.Cumulative frequency is the ratio that the cumulative frequencies of this section of learning state data accounts for all section cumulative frequencies summations, corresponding is time ratio, it it is the persistent period length of this section of learning state data ratio that accounts for all section learning state data acquisition time overall lengths, in the case of brain electric transducer acquisition time interval is constant, it is long that frequency and the product of acquisition interval are exactly the time, i.e. frequency ratio is identical with time ratio, frequency ratio or time ratio reflect the proportion that this segment data accounts in total data, more intuitively reflection learning state data distribution characteristics.
The flowing water form of grouped record is:
Grouping feature values 1, this section of persistent value
Grouping feature values 2, this section of persistent value
Grouping feature values 1, this section of persistent value
……
Continuous-flow type recording feature value and persistent value, operation of recording is simple, and subsequent treatment can be carried out with combining also.In this recording method, eigenvalue can be the fragmentation value set, or this segment data statistical value, the size of statistical value is in packet section, and the statistical value in same packet section can equal can not also wait, and the information content of its reflection is in further detail, further improve, can record packet zone segment value, this section of statistical value, this section of persistent value, the statistical value of same packet section can further be analyzed and processed, meet the needs of different levels by subsequent treatment.
Section value arrange can be one to multiple, wherein, for multiple fragmentation values, can be placed equidistant, or different away from setting, reduce the spacing of emphasis monitoring part, amplify remainder, accomplish that weight is close light thin, or take interruption method, secondary sections segment data is interrupted dispensing.
Carrying out segmentation according to waveform parameter, waveform is the geometry of learning state data midbrain wavelength-division cloth, and the learning state of special state has special geometry, as reflected dormant spindle wave.Waveform parameter analysis is to carry out matched jamming according to wave character, and the learning state data of waveform of the same race are divided into same section, such as learning state data spindle waveform occur is classified as section of sleeping.Wave character mainly extracts geometric properties by the method for time-domain analysis, such as the analysis of zero passage section, histogram analysis, variance analysis, coherence analysis etc..It is combined value according to the section value that waveform parameter is arranged, waveform parameter includes amplitude, average, variance, skewed degree, kurtosis etc., different waveforms has different parameters to combine, learning state geometry is divided into coupling or does not mates, calculate the statistical value of learning state data of coupling as eigenvalue, or using the corresponding mental status as eigenvalue, using duration or cumulative frequencies as persistent value, or record time started and end time.
(2) classify according to the mental status in learning state data, it is directly to carry out section partition by the mental status, learning state data under the same mental status are divided into one section, such as according to sleepy, loosen, tranquil, be absorbed in, the criterion of excited five kinds of mental status divides five sections, learning state data under sleepy are divided into one section, adding up persistent value and the meansigma methods of this segment data, record, under " sleepy " section, records data under other state equally.Judge that a kind of method of the mental status is according to the brain wave frequency range prestored or brain wave data mark sheet, it is judged that the affiliated mental status, if brain states data match with set point or feature, it may be determined that the current mental status is classified.In addition, judge that the mental status can also be calculated by neural algorithm to get, i.e. calculate according to brain wave frequency spectrum data and set up criterion, the mental status that different neural algorithms calculates has deviation, generally Binding experiment data are carried out, and the criterion of foundation can be single setting threshold values, it is also possible to by the compound condition of multiple combinations of values, the brain wave data of algorithms of different foundation is the most different, and mental status determination methods can consult correlation technique data.
2, Time segments division
Described Time segments division includes that the relative size according to learning state data carries out segmentation, fluctuation set point one continuous time section learning state data be divided into one section, or carry out slot setup according to the time, it is divided into one section being in the learning state data setting the time period.The Time segments division carried out according to learning state data and the Time segments division carried out according to time conditions, it it is all record the most successively, but the emphasis of both reflections is different, the Time segments division carried out according to learning state data, reflect is size and the length of persistence of the learning state data of each period, its persistent value and eigenvalue are all being continually changing, need to calculate and extract, the Time segments division carried out according to time conditions, its eigenvalue is continually changing, need to calculate and extract, and persistent value is to set, stress reflection is the size of the learning state data of each period.
(1) Time segments division is carried out according to learning state data relative size, it is that learning state data close for size are divided into one section, using this period learning state data statistics value as eigenvalue, fragmentation value is set by learning state data difference or ratio or dispersion, fluctuation is divided into one section in the learning state data of a continuous time of set point.
Learning state data difference is to compare the distance size between two learning state data, the fragmentation value arranged according to differential technique, at a distance of being divided into same section less than the learning state data of fragmentation value, i.e. fluctuation change is divided into same section in the learning state data of certain amplitude scope, when difference is more than fragmentation value, illustrate that fluctuation becomes big, go beyond the scope, previous data are classified as present period, up-to-date data are classified as subsequent period statistics, the time separation of two sections is the acquisition time of latest data, or this time point carries previous acquisition time and is spaced as separation.The computational methods of difference have position differential technique and central difference method.Position difference is to be ranked up learning state data by ascending order or descending, calculate the difference of two specific tagmeme numerical value, i.e. N bit data and the absolute deviation of N bit data reciprocal, the difference of such as maxima and minima or upper quartile and lower quartile absolute deviation, what the fragmentation value arranged according to position difference reflected is the height scope of data fluctuations, when it is less than setting fragmentation value, illustrates that numerical values recited is close, be classified as the same period.Central difference is the difference calculating study status data with center value, center value includes meansigma methods, standard deviation, any one in median and mode, learning state data and the difference of center value, reflect the off-centered distance of these data, central difference is divided into potential difference (more than the difference of the data of central value Yu center value) and lower potential difference (less than data and the difference absolute value of center value of central value), what the fragmentation value set according to central difference reflected is the upper half range of data fluctuations or lower half range scope, the judgement of central difference uses upper and lower potential difference simultaneously less than setting district segment value.Center value can first pass through position difference and judge to set up, and when the data amount check of same section is more than N, calculates center value, it is proposed that N is not less than 6.After center value is set up, can keep constant, it is also possible to the statistical updating according to the increase of this segment data, or converting new learning state data adds up again.
The reflection of learning state data ratio is data degree of deviations, including position ratio and center ratio.Position ratio is the ratio calculating two specific tagmeme numerical value, or calculate the difference of two specific tagmeme numerical value, again divided by any value in center value or upper end position numerical value or lower end position numerical value, center ratio is to calculate study status data and the ratio of center value, or after calculating study status data and central difference, then divided by center value.The fragmentation value set according to ratio, is that described ratio is classified as the same period less than the data of setting value, and concrete grammar is with difference segmentation method.
The diversity of dispersion reflection learning state data, is to pass judgment on the excentric trend of learning state data, is to estimate data individual character, and dispersion includes any one in different many ratio, quartile deviation, mean deviation, standard deviation, coefficient of dispersion, standardized value.Dispersion, as a kind of statistical value, should have N number of data statistics to draw, it is proposed that N is not less than 6.The fragmentation value arranged according to dispersion is one and sets threshold values, if the dispersion of batch of data is less than setting threshold values, illustrate that this batch data has stronger general character, this batch data is divided into same section, it is continued for statistics to go down, once dispersion is more than setting threshold values, illustrate that the diversity of data becomes big, it is divided into subsequent period making dispersion become big latest data, data before these data are divided into the same period, calculate the statistical value of this one piece of data not including latest data as eigenvalue, as the persistent period length of persistent value or the end point of cumulative frequencies also at latest data, the acquisition time of these data is end time this period point and new period start time point.Standard deviation is a key value in dispersion calculates, when sample size is bigger, the impact on standard deviation of subsequent samples can weaken, it is proposed that control the sample size of standard deviation, i.e. the data re-segmenting to same period statistical standard difference according to learning state market demand feature.
Carrying out Time segments division according to learning state data, the learning state data feature values of each period is statistical value or the mental status classification of this segment data, using duration or cumulative frequencies as persistent value, record format is:
Statistical value 1, this period lasts value 1
Statistical value 2, this period lasts value 2
Statistical value 3, this period lasts value 3
……
By a fragmentation value set, can mark off multiple period data, the statistical value of each period data constantly fluctuates change, and above-mentioned statistical value 1 is not equal to statistical value 2, but can be equal to statistical value 3.Further refinement, can reflect more rich information content with minute book period start time or end time in record.
(2) Time segments division is carried out according to time conditions, it is to carry out slot setup according to the time, the one section of successive learning status data meeting time conditions is divided into one section, record the statistical value of this period status data, or further the learning state setting the time period is carried out section partition, the status data eigenvalue of each section in record sets the period successively and persistent value.Such as a example by student attends class scene, divide according to the classroom timetable setting period, record the study condition of every class respectively, it is to avoid the record ignored of break inactivity period；Or hall class process is temporally carried out segmentation record, such as set every 1 or 3 or 5 minute as segment record when one once, using the learning state data statistics value of each period as eigenvalue, segment length or cumulative frequency when persistent value is for setting, it is also possible to record each period start time；Further refinement, the learning state data of hall class hour section can also be carried out section partition, the status data of same section next one continuous time is divided into a period, records status data eigenvalue and the persistent value of different periods, the state feature during reflection classroom learning successively.By the data of Time segments division record, can reflect that the learning state data of user every class, along with the situation of change of class period, reproduce classroom learning process, the learning state of the different course of contrast, and break inactivity period interruption is dispensed.
Time segments division is carried out according to the time, if the period is a setting value, i.e. the long record of timing, record data can be omitted persistent value, only record statistical value, by the corresponding described period of arranging in order:
Statistical value 1
Statistical value 2
Statistical value 3
……
If the period is multiple setting values, i.e. become duration record, segmentation duration is different, segment length time some, some periods are short, then can record the duration of each period and be distinguish between, if Time segments division is discontinuous, there is omission break time, can be segmentation statistical value record under day part, record format is: period, statistical value.
The aforementioned segmentation method according to learning state data is individual data and the comparing in real time of fragmentation value, judge segmentation belonging to each data one by one, for avoiding the impact of abnormal data, judgement can be compared according to the meansigma methods of a collection of M learning state data, when meansigma methods meets segmentation requirement, illustrate that these M data is close, same section can be classified as, in order to keep the seriality judged, after a new data occurs, the first of these M data data is classified as this section of statistics, continue to remain M data, calculate its meansigma methods to judge, if meansigma methods is unsatisfactory for segmentation requirement, illustrate that the new data occurred is the data of another section, last data is classified as next section, again other data are all classified as the last period, and add up the persistent value of this section.
Another kind of many data judgment method are when there being N number of data to meet segmentation requirement in M learning state data, these M data is classified as same section, in order to keep the seriality judged, after a new data occurs, the first of these M data data is classified as this section of statistics, continue to remain M data, when meeting the data amount check of segmentation requirement less than N number of, illustrate that the new data occurred is the data of another section, last data is classified as next section, again other data are all classified as the last period, and add up the persistent value of this section.Above-mentioned M data are continuous data, and N number of data can be continuously or discontinuously data.Identical method can also be according to meeting the temporal summation that segmentation requires and judge setting in duration.
The segmentation record of learning state data, all data that memorizer can gather this are placed on a file, different segment data records in the different files that this document presss from both sides, are distinguish between by file name, such as using eigenvalue as file name；Or in one file of all data records that this is gathered, different segment datas is distinguished by paragraph position, or add segmentation markers according to segmentation result, the segmentation markers of same section value is identical, the segmentation markers of different section values is different, during follow-up data Treatment Analysis, by identifying that segmentation markers sub-elects required segment data.
Preferably, recording equipment can also arrange prompting module, when learning state data fall into a certain setting section, processor triggers prompting module and reminds, the fragmentation value such as set according to head pose, select skew angle section bigger than normal or a skew angle is set reminds threshold values, when head pose enters this section or more than this threshold values, processor triggers prompting module and reminds；Or remind apart from section from the fragmentation value set, setting difference according to eye distance, and remind section to take difference based reminding method according to difference, such as closer distance section is slightly reminded, and closely section is slightly brought up again awake, close distance section severe prompting etc..Additionally, can also fall asleep in excited several mental status, select certain mental status, if relaxation state, sleepy state, the fragmentation value of hypnagogic state or criterion are as reminding threshold values, when described learning state data value falls into this segment limit or presents the bad mental status, processor triggers prompting module and reminds, and use different prompting degree or mode according to the different mental status, somewhat remind under such as relaxation state, slightly bring up again awake under sleepy state, under hypnagogic state, carry out severe prompting.As a example by LED in institute's diagram, somewhat remind carry out slow flash sparkle, slightly bringing up again wake up flashes soon, severe remind persistently become clear.
In the configuration block diagram of display device shown in Fig. 6, including the communicator receiving learning state data, data are carried out processor and the display of display statistical result of segmentation statistics, the learning state data that communicator is received by processor carry out sectional statistics, eigenvalue and persistent value according to each section of learning state data draw statistical graph, transmit display and show.
Display device can be smart mobile phone or other there is data receiver and process the equipment integrating of display function, or formed by communicator, processor, the Split assembled of display.
The learning state data that communicator reception/recording device transmits, can be a kind of data transmission device or interface, or have the equipment of data receiver function, as smart mobile phone, the webserver, computer and other there is the equipment of data receiver function.
Communicator can obtain learning state data by wired mode, as received data by USB USB (universal serial bus).
Communicator can receive learning state data by SD storage card etc. as transmission medium.
Communicator can receive learning state data wirelessly, such as any one in bluetooth, radio frequency, infrared ray, zigbee, wifi；Or receive data by the Internet, LAN.
Processor includes processing unit such as CPU and the memory element of storage program carrying out processing controls, processor is as the ingredient of display device, display device can be independently of exist, such as server, after data are processed, result is sent to mobile phone show, or processes the smart mobile phone of display one, directly show after smart mobile phone processes.If smart mobile phone is as complete display device, communicator, processor, display become the Inner Constitution unit of display device.
Display is the terminal unit with menu display function, can integrate with communicator and processor, it is also possible to individualism, such as mobile phone, computer and have the miscellaneous equipment of display function and can serve as the stand alone display of display device.
The segmentation of learning state data shows, first according to above-mentioned segmentation method, learning state data are carried out segmentation, data form after process has continuous-flow type and statistics formula two kinds, continuous-flow type is in chronological order, one segmentation one record, the section of same eigenvalue can repeat, and statistics formula is to remerge continuous-flow type, and the section of an eigenvalue only records accumulative total persistent value.For the segment data of continuous-flow type, statistical graph can be according to the time sequencing of segmentation, and with the time as transverse axis, persistent value draws broken line graph or rectangular histogram for the longitudinal axis or block diagram show, or working out the period form corresponding with persistent value shows.The chart of statistics form with section value as transverse axis, with persistent value for the longitudinal axis, can be drawn broken line graph or rectangular histogram according to the size order of piecewise interval or block diagram shows, or the form working out section value corresponding with persistent value shows.The reflection of continuous-flow type chart is the learning state data concrete situations at different periods, and what statistics formula chart reflected is the population distribution feature of learning state data.Mental status statistic histogram shown in Fig. 7, the longitudinal axis is cumulative frequency, and transverse axis is section value, the most corresponding sleepy, loosen, tranquil, be absorbed in, excited five kinds of mental status.
Claims (10)
1. a learning state fragmented recording strategy, its feature comprises the following steps:
(1) processor carries out segmentation according to imposing a condition to learning state；
(2) extract this section of status data length of reflection or the persistent value of period, or extract this section of status data size of reflection or the eigenvalue of state；
(3) eigenvalue of minute book section learning state data and persistent value.
Recording method the most according to claim 1, it is characterized in that: described segmentation includes learning state is carried out section partition, the status data of same section is divided into one section, or learning state is carried out Time segments division, one continuous time section status data be divided into one section.
Recording method the most according to claim 2, it is characterized in that: described section partition includes arranging fragmentation value according to learning state data, the status data falling into same section is divided into one section, or classify according to the mental status in learning state, the status data under the same mental status is divided into one section.
Recording method the most according to claim 3, it is characterized in that: described any one numerical value that fragmentation value includes in the size according to learning state data or waveform parameter be set according to learning state data fragmentation value is set, the status data of same section is divided into one section.
Recording method the most according to claim 2, it is characterized in that: described Time segments division includes that the relative size according to learning state data carries out segmentation, fluctuation set section one continuous time section status data be divided into a period, or carry out slot setup according to the time, it is divided into a period being in the learning state data setting the time period.
Recording method the most according to claim 5, it is characterized in that: described setting section is that any one numerical value in the difference according to learning state data or ratio or dispersion arranges fragmentation value, meet segmentation condition one continuous time section status data be divided into one section.
Recording method the most according to claim 1, is characterized in that: described persistent value includes that the cumulative frequencies of described section of learning state data or cumulative frequency or persistent period length or persistent period length account for any one in the period of the ratio of total time length or this period of time started and end time or described period；Described eigenvalue includes any one in described section of learning state data sectional value or described section of learning state data statistics value or described section of mental status classification.
8. a learning state segmentation recording equipment, including wear-type framework, learning state sensor, processor and the memorizer being placed on framework, learning state sensor acquisition learning state data transfer processor carries out calculating process, transmit memory recording address book stored, it is characterized in that:
(1) processor performs method step described in claim 1 ~ 7, and learning state is carried out segmentation；
(2) extract this section of status data persistence length of reflection or the persistent value of period, or extract this section of status data size of reflection or the eigenvalue of state；
(3) eigenvalue of memorizer minute book section learning state data and persistent value.
Recording equipment the most according to claim 8, is characterized in that: described learning state sensor includes head pose sensor, uses at least one in eye range sensor, brain electric transducer.
10. learning state is according to a segment display device, including the communicator of reception learning state data, data carries out processor and the display of display statistical result of segmentation statistics, it is characterized in that:
(1) processor performs the method step described in claim 1 ~ 7, and the learning state data receiving communicator carry out segmentation；
(2) extract reflection this segment data length or the persistent value of period, or extract this section of status data size of reflection or the eigenvalue of state；
(3) processor draws the eigenvalue of each section of brain wave data and the statistical graph of persistent value, transmits display；
(4) display shows described statistical graph.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510111008 | 2015-03-13 | ||
| CN201510111008X | 2015-03-13 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105962930A true CN105962930A (en) | 2016-09-28 |
Family
ID=56989000
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610119931.2A Pending CN105962930A (en) | 2015-03-13 | 2016-03-03 | Learning state sectional-type recording method and device, and learning state sectional-type displaying device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105962930A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107292786A (en) * | 2017-07-13 | 2017-10-24 | 广东小天才科技有限公司 | A kind of learning time statistical method, device and terminal device |
| CN108836323A (en) * | 2018-05-08 | 2018-11-20 | 河南省安信科技发展有限公司 | A kind of learning state monitoring system and its application method based on brain wave analysis |
| CN110840430A (en) * | 2018-08-21 | 2020-02-28 | 北京万生人和科技有限公司 | Intra-abdominal pressure data screening method, computer-readable storage medium, and intra-abdominal pressure data screening device |
-
2016
- 2016-03-03 CN CN201610119931.2A patent/CN105962930A/en active Pending
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107292786A (en) * | 2017-07-13 | 2017-10-24 | 广东小天才科技有限公司 | A kind of learning time statistical method, device and terminal device |
| CN108836323A (en) * | 2018-05-08 | 2018-11-20 | 河南省安信科技发展有限公司 | A kind of learning state monitoring system and its application method based on brain wave analysis |
| CN108836323B (en) * | 2018-05-08 | 2021-01-12 | 河南省安信科技发展有限公司 | Learning state monitoring system based on electroencephalogram analysis and using method thereof |
| CN110840430A (en) * | 2018-08-21 | 2020-02-28 | 北京万生人和科技有限公司 | Intra-abdominal pressure data screening method, computer-readable storage medium, and intra-abdominal pressure data screening device |
| CN110840430B (en) * | 2018-08-21 | 2022-09-13 | 北京万生人和科技有限公司 | Intra-abdominal pressure data screening method, computer-readable storage medium, and intra-abdominal pressure data screening device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105962929A (en) | Brain wave data sectional-type recording method and device, and brain wave data sectional-type displaying device | |
| CN105962931A (en) | Learning scene sectional-type recording method and device | |
| US10517527B2 (en) | Sleep quality scoring and improvement | |
| CN104720748B (en) | A kind of sleep stage determines method and system | |
| JP4825586B2 (en) | Sleep monitor device | |
| CN107106085A (en) | Apparatus and method for sleep monitor | |
| US20070249953A1 (en) | Method and apparatus for detection of nervous system disorders | |
| CN110491468A (en) | A kind of processing method and system of sleep quality report | |
| CN109215791A (en) | Health control method, system, equipment and storage medium based on multi information decision | |
| CN109222950B (en) | Data processing method and device | |
| EP2846683A2 (en) | System and method for monitoring the health of a user | |
| KR102100120B1 (en) | Method, apparatus and computer program for monitoring of bio signals | |
| CN104720746A (en) | Sleeping stage determination method and system | |
| KR102383921B1 (en) | Methods and system for customized sleep management | |
| CN105962930A (en) | Learning state sectional-type recording method and device, and learning state sectional-type displaying device | |
| CN107343786A (en) | Based on the packet of state of mind brain wave data and statistical method and record display device | |
| CN111657855B (en) | Sleep evaluation and sleep awakening method and device and electronic equipment | |
| CN115862873B (en) | Method, system and device for quantifying and intervening sleep rhythm | |
| CN104615851A (en) | Sleep monitoring method and terminal | |
| CN105976579B (en) | Bad learning state based reminding method and alarm set | |
| CN104407768A (en) | Terminal | |
| US20160240098A1 (en) | Smart tablet-based neurofeedback device combined with cognitive training, method and computer-readable medium thereof | |
| US11033214B1 (en) | Wearable eye tracking system | |
| CN109922718A (en) | For providing the method, apparatus and computer program product of dynamic wake-up alarm | |
| CN113261929B (en) | Mobile phone bad use behavior risk early warning system based on heart rate variability index |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20160928 |